PP1310 - Effect of Cochlear Implant Electrode Array Insertion Depth on Bimodal Benefit in Speech Perception

Shallower insertion depths of cochlear implant (CI) electrodes create frequency-to-place mismatch (i.e., spectral mismatch). Different degrees of residual hearing in the hearing aid (HA) ear interact with spectral mismatch in the CI ear. In this study, using acoustic simulation of bimodal hearing, we determined the spectral mismatch range that results in a significant decline in speech perception for different degrees of residual hearing. Results indicated a significant decline in speech perception when spectral mismatch was greater than 4 mm. Spectral mismatch was not a significant factor for individuals with residual hearing up to 2000 Hz.


Summary:

Rationale/Purpose

Bimodal hearing, a cochlear implant (CI) in conjunction with a hearing aid (HA) on the opposite ear, produces a considerable synergistic effect. While many bimodal users experience benefits, others gain little to none. Two potential contributors to this variability are different degrees of residual hearing thresholds in the HA ears and their interactions with different insertion depths of the CI electrodes, which result in a spectral mismatch within the CI ear. The spectral mismatch is known to be a detrimental factor for speech perception. In this study, using simulations of bimodal hearing with normal hearing (NH) listeners, we determined the spectral mismatch range that results in a significant decline in speech perception for different degrees of residual hearing.

Method

Nineteen adult NH listeners were recruited for speech perception as a function of spectral mismatch and residual hearing. Three acoustic hearing losses were created using band-pass filters (50-500, 50-1000, or 50-2000 Hz). For electric simulation, an 8-channel sinewave vocoder was used with a fixed output frequency range (621-7000 Hz) which corresponds to a 24 mm insertion depth and seven different input frequency ranges to create seven different insertion depths in mm: 18, 20, 22, 24, 26, 28, and 30. As a result, relative to the 24 mm output, there are seven spectral mismatches in mm: -6, -4, -2, 0, +2, +4, and +6. Speech perception was measured in quiet and noise at 0- and 6-dB signal-to-noise ratio (SNR) with HA alone to the left ear, CI alone to the right ear, and three bimodal conditions (i.e., CI+500Hz, CI+1000Hz, and CI+2000Hz). For the bimodal conditions, the HA alone signal was presented to the left ear, while the CI alone signal was presented to the right ear. 

Results and Conclusions

The speech perception for HA alone depends on the residual hearing and the SNR, as exemplified in A2000 yielding the best results. For the CI alone condition, the spectral mismatch range broadens with increasing SNRs. Speech performance was considerably declined with 2, 4, and 6 mm spectral mismatch at SNRs of 0 dB, 6 dB, and in quiet, respectively. For the CI+500Hz, the spectral mismatch range is consistent regardless of SNRs. Speech performance considerably declined when spectral mismatch was greater than 2 mm. For the CI+1000Hz, the spectral mismatch is not a factor in quiet, while the spectral mismatch range is 6 mm in noise. For the CI+2000Hz, the spectral mismatch range becomes almost flat across SNRs. Bimodal hearing is negatively affected when CI spectral mismatch is greater than 2 mm in noise and 4 to 6 mm in quiet. Better residual hearing helps to reduce the negative effect of CI spectral mismatch, leading to better speech perception. These results are clinically important as frequency mapping can be created to avoid a significant decline in speech perception by mapping frequency ranges within the spectral mismatch range. Results show that the spectral mismatch range can be listener-specific, making aural rehabilitation treatment goals highly individualized, leading to better clinical recommendations.